sign in sign up
<<
Home , Axial Seamount, Cobb hotspot

University of South Florida Scholar Commons

Marine Science Faculty Publications College of Marine Science

9-1984

Sr Isotopic Variations along the Juan de Fuca

Jacqueline Eaby U.S. Geological Survey, [email protected]

David A. Clague U.S. Geological Survey

John R. Delaney University of

Follow this and additional works at: https://scholarcommons.usf.edu/msc_facpub

Part of the Life Sciences Commons

Scholar Commons Citation Eaby, Jacqueline; Clague, David A.; and Delaney, John R., "Sr Isotopic Variations along the " (1984). Marine Science Faculty Publications. 1321. https://scholarcommons.usf.edu/msc_facpub/1321

This Article is brought to you for free and open access by the College of Marine Science at Scholar Commons. It has been accepted for inclusion in Marine Science Faculty Publications by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 89, NO. B9, PAGES 7883-7890, SEPTEMBER 10, 1984

Sr IsotopicVariations Along the Juande Fuca Ridge

JACQUELINEEABY AND DAVID A. CLAGUE

U.S. GeologicalSurvey, Menlo Park, California

JOHN R. DELANEY

OceanographyDepartment, University of Washington,Seattle

Srisotopic ratios of 39glass and microcrystalline samples along the Juan de Fuca Ridge and 1 glasssample from BrownBear are at the lowerend of the rangefor normalmid-oceanic ridge basalt(MORB)' the average87Sr/S6Sr ratio is 0.70249+ 0.00014(2-a). Althoughsubtle variations exist alongstrike of the ridge,the Sr isotopedata do not showsystematic variation relative to the proposed CobbHotspot. The isotopicdata are inconsistentwith an enrichedmantle-plume origin for the Cobb- EickelbergSeamount chain, as has beenproposed for Iceland,the Azores,and the Galapagosspreading center.Sr isotopicratios of samplescollected north and southof the Cobb offsetare identical,although minor elementratios indicate that theseregions have chemicallydistinct sources. These distinct mantlesources may not havebeen separated long enough to developisotopic differences.

INTRODUCTION characterizedby high 2ø6pb/2ø'*Pb,low 87Sr/86Sr,and high The hotspo•tmodel proposed by Wilson[1965] and Morgan •'•3Nd/•'•'•Nd.Geochemical variations of lavasalong / [!973] explainsthe origin of linear seamountchains as the ridge systemsare interpretedto be the result of mixing be- result of movementof lithosphericplates over mantle plumes. tween enriched and normal MORB end-members [Schillin•t, The Cobb-EickelbergSeamount chain, in the northeastPaci- 1975a; Schillin•tet al., 1982a]. fic, trends northwestward from the Juan de Fuca Ridge and is The objectiveof thisstudy is to presentnew Sr isotopicdata interpretedto have a hotspot origin. A seamountat approxi- for dredgedfrom the Juan de Fuca Ridge and to use these data to evaluate the mantle source characteristics of the mately lat 46øN on the axis of the Juan de Fuca Ridge ()is coincidentwith the intersectionof the Cobb- . EickelbergSeamount chain and the ridge; it is thought to be GEOLOGIC SETTING the site of presenthotspot activity, herein referred to as the CobbHotspot [Vogt and Johnson, 1975]. The Juande Fuca Ridge,which extends 500 km from the Detailed studiesof other mid-oceanicridge/hotspot systems, BlancoFracture Zone to the SovancoFracture Zone (Figure suchas Iceland [Hart et al., 1973; $chillin•h1973, 1975a], the 1),is a spreadingcenter of moderate rate, 30 •ma-x between Azores [White et el., 1975, 1976; $chillin•h 1975b], and the the Pacificand Juan de Fucaplates. Mor•tan [1973] proposed Galapagosspreading center [$chillin•l et el., 1982a;Verma and theexistence of a •otspoton theJuan de FucaRidge to Schilling,1982] showed that basalt erupted along deep mid- explainthe northwest trending linear seamount chain includ- oceanicridge segments(normal MORB) differs from basalt ingBrown Bear, Cobb, and Eickelberg . K-Ar ages erupted along topographic highs associatedwith islands fromHorton Seamount, located near the northwest end of the astride the ridge (enriched MORB). Normal MORB is seamountchain, indicate that the Cobb Hotspot has been a characterizedby low concentrationsof incompatibleelements; zoneof high volcanicproductivity for at least20 Ma [Turner light rare elementdepleted patterns; high K/Ba, K/Rb, et el., 1980]. and Zr/Nb ratios; low 8•Sr/S6Sr ratios; and high There are four morphologicallydistinct segmentsof the •'•3Nd/X'•'•Ndratios [White et el., 1975;DePaolo and Wass. Juande Fuca Ridge(Figure 2)'(1) A 90-km-long,linear, sym- erburg,1976]. These chemical characteristics are consistentmetrical southern ridge segmentnorth of the BlancoFracture with those of basalt derived from mantle sourceregions deple- Zone;(2) a 140-km-longcomplexly disturbed central segment, ted by previousmelting events[Kay et el., 1970]. Enriched recentlyoffset approximately 20 km westof themagnetic axis MORB, which has chemical affinities to oceanic island tho- of symmetry,which includes the axial seamount;(3) another leiites, differs from normal MORB by having higher con- 90-km-long,linear, symmetricalnorthern ridge segmentex- centrationsof incompatible elements;rare earth element pat- tendingnorthward to the Cobboffset (the tip of a northward ternsmore enriched in thelight elements; lo•wer K/Ba, K/Rb, propagatingrift [Johnsonetal., 1983]); and (4) the ridge north and Zr/Nb ratios; higher 8*Sr/8aSr ratios; and lower ofthe Cobb offset, referred to asthe Endeavor Ridge [Delaney x'•3Nd/X'•'•Ndratios than normalMORB [Bryanet el., 1976; et al., 1982]. White and Bryan, 1977; O'Nions et el., 1977]. Schilling[1973, Basalt collectedsouth of the Cobb offset (regions 1-3) is 1975b] and Schillinget el. [1982a] proposethat enriched predominan,tlyaphyric olivine- to quartz-normativetholeiite MORB is generatedfrom a nondepletedmantle source.Zind- slightlyto distinctlyenriched in iron and titaniumcompared let et el. [1982] propose a third chemically independent with normalMORB. Regionalvariations of FeO* (total iron mantle component,possibly related to recycledoceanic crust, calculatedas FeO) are plotted as a functionof latitude in Figure3. Ferrobasalt {> 10% FeO*) is particularly abundant Thispaper is not subjectto U.S.copyright. Published in 1984by at the southernend of region 1, and 20-30 km southof the the AmericanGeophysical Union. Cobb offset. The maximum iron and titanium enrichment in Paper nurlaber4B0708. the lavasoccurs behind zones identified as propagatingrifts in

7883 7884 EABY ET AL.' SR ISOTOPIC RATIOS ALONG JUAN DE FUCA RIDGE

chemistry,and glasschemistry of a muchlarger set of samples will be discussedelsewhere (J. R. Delaney et al., unpublished data, 1984).All samplesare glassybasalt from the actively spreadingaxis and werechosen for their fresh•:essand spatial distribution.Samples along the ridgeand from the peak of the • I -• WelkerGuyol axial seamountare approximately25 km apart. ß • ß , •t • ANALYTICAL TECHNIQUES .- . •. ,' ', - • Hand-picked,fresh glasschips were rinsedin acetoneand __ ' • • ' ' ' •m,Chain • 'O DIlwood . washed with distilled water in an ultrasonic cleaner for 5 min. Microcrystalline-basaltsamples were ground to < 200 meshin .... , -- .'• i • a tungstencarbide shatterbox.Samples of powder or glass I. '/ ....,, , •o•LL/• :5 • 'a weighingapproximately 100 mg were dissolvedin HF and HC10,•, and Sr was separatedfrom the unspikedaliquots

i i '1 '1 usingstandard ion-exchange chromatography. Mass analysis was done at the U.S. Geological Surveyin Menlo Park on a

/ . ' ' "•,• •'Y

• • I • • 40øN / I '' I 1 •1 •1 I •.,..:S'o,."?.•I I 49 ø N 150ø 1400 I •0 ø 1200W Fig. 1. Seamount and ridge distribution in the northeast Pacific Ocean. The youngest seamount of the Cobb-Eickelberg Seamount chain occurs at the intersection of the seamount chain and the Juan de Fuca Ridge. Figure modified from Barr [ 1974]. %0 ENDEAVOUR . both regions [Hey and Wilson, 1982; Delaney et al., 1981; SEGMENT - 48øN Johnsonet al., 1983]. The petrologic diversity of lavas in re- (region 4) gions 1-3 may be the result of variable degreesof fractional crystallization[Clague and Bunch,1976] in chambers spatiallyassociated with propagatingrifts [Sintonet al., 1983]. COBB OFFSE'T Vogt and Byefly [1976] postulated that the iron and titanium enrichmentcould result from fractional crystallizationin a 47øN NORTHERN subaxial conduit transporting magma longitudinally away SEGMENT from the hotspot. (region 3) Basalt from region 4 is chemicallydistinct from basalt south of the Cobb offset.The lavas are not unusuallyenriched in iron and titanium but are enrichedin K20 (up to 0.56 wt %), Na20 (up to 3.36 wt %), and SiO: (up to 52.5 wt %). The CENTRAL P:Os/K:O ratios rangefrom 0.5 to 0.8 comparedwith P:O5 SEGMENT 46øN /K20 ratios of > 1.0 for lavas south of the Cobb offset (region 2) (J.R. Delaney et al., unpublisheddata, 1984).The P:Os/K:O ratios are probably indicative of mantle source regions [Clagueet al., 1981-1,thus suggestingthat from region4 is 4R derived from mantle sources distinct from those of lava from regions 1-3. 46 45 øN SAMPLE SELECTION A comprehensivesampling and surveyingprogram of the 4 500-km-long ridge segmentbetween the Sovancoand Blanco I 'if3 i n•q t2(1 124 fracturezones was initiated by the Universityof Washington in 1980 to characterizethe chemical,petrologic, and tectonic interaction among a spreading center of moderate rate, a I 44øN 131 o ',1/ 128oW 127ø w seamount-generatinghotspot, and propagatingrifts. Ratios of 87Sr/86Srwere determined on 39 glassand microcrystalline- Fig. 2. Locations of dredges sampled for this study. The + 200 basalt samplesfrom 30 locations along the Juan de Fuca and + 500 7 magnetic anomaly contours shown are from National Ridge and one sample from the south slope of Brown Bear Ocean Survey magneticanomaly map. Symbolsfor dredgesare solid Seamount. Samples were selected from dredge hauls from circles,University of Washington cruise TT152; solid squares,Uni- versity of Washington cruise TT170; open square, U.S. Geological cruises152 and 170 of the R/V T. Thompson(University of Survey cruise L1 i81WF; solid triangles, State University Washington)and cruise81-017 of the R/V Hudson(University cruise Y74-1; and open circles,University of British Colombia cruise of British Columbia).Additional sampleswere obtainedfrom HU81-017. Forty samples from 32 dredges were analyzed for cruise74-1 of the R/V Yaquina(Oregon State University)and 87Sr/S6Srratios in this study.Also shownare the divisionsbetween the four morphologicallydistinct ridge segments,the Cobb offset,and cruiseL 1181WF of the R/V S.P. Lee (U.S. GeologicalSurvey). the bathymetric outlines of Cobb and Brown Bear seamountsand the Microprobe analysesof the glasssamples analyzed for Sr iso- axial seamount (see text for discussion).Figure modified from J. R. topes are given in Table 1. Detailed petrography, mineral Delaney et al. (unpublisheddata, 1984). EABY ET AL.: SR ISOTOPIC RATIOS ALONG JUAN DE FUCA RIDGE 7885

0.70249 ñ 0.00014 (2-a). There are subtle variations along strike, but they do not correlate with tectonic or petrologic ß variations.The resultsof a probability of statisticaldifference (F(z)) calculation are shown in Table 3. The variations in 87Sr/S6Srratios described below were distinguished by eye. The Sr isotopicdata do not show systematicvariation as- ee 'e• ,"• ßNormal MORB sociatedwith the Cobb Hotspot. The 13 samplesthat spatially bracket the axial seamount from latitude 45 ø to 46ø10'N I (group b) have essentiallyuniform Sr isotopicratios with an lOOO averagevalue of 0.70250ñ 0.00010 (2-a). Sampleswith the lowest8•Sr/86Sr ratios, 0.70235 ñ 0.00006,are locatedon the 2000 northern flank of the axial seamount (group c) and are some 3000 of the lowest ratios reported for any MORB [White and Schilling,1978].

o o 417ø The 87Sr/86Srratios of samplesat the southernend of SOUTHERN CENTRALNORTHEl::•NI EN4D•VOUR l region 1, from the to latitude 45øN (group a), are slightly higher than the rest of the ridge; their averagevalue is 0.70260 ñ 0.00009. This relative maximum is coincidentwith a high in iron and titanium enrichment.Hedge SOVANCO and Peterman [1970] reported an analysisof 0.7025 ñ 0.0003 BLANCO FRACTURE ZONE FRACTURE ZONE for a samplefrom thisregio n. The Sr isotopicratios df samplesnorth and south of the Fig. 3. Regional variation of FeO* content and for Cobb offset (regions4 and 3' group d) are statisticallyidenti- the Juan de Fuca Ridge. Stippled field surroundscomprehensive data set (J. R. Delaney et al., unpublisheddata, 1984); FeO* of samples cal and have an average value of 0.70246 ñ 0.00009. The two analyzedfor 87Sr/86Srin this studyare shownas solidcircles. Bath- analysesreported by Hedgeand Peterman [1970] for samples metry from the actively spreadingaxis shows the axial seamount at from this area have 87Sr/86Sr ratios of 0.7022 and approximately46 ø N that is thought to be the most recent activity of 0.7025 + 0.0003 and are in agreement with values from this the hotspot that producedthe Cobb-EickelbergSeamount chain. study. The Sr isotopic data of lavas from Brown Bear and Cobb doublecollector MAT 261 automatedspectrometer utilizing a seamountssuggest that both normal and enriched type lavas double-filament source. (Any use of trade names and trade- are present.A tholeiite sample from the south slope of Brown marks in this report is for descriptivepurposes only and does BearSeamount has a 87Sr/86Srratio of 0.70251ñ 0.00006and not constituteendorsement by the U.S. Geological Survey.) isotopically resemblesthe ridge tholeiites. Subbarao et al. Data were collected in 26 mass scans of 85-86-87-88. The data [i973] reported87Sr/86Sr ratios of threesamples from Cobb werenormalized to a 87Sr/Sa•rratio of 0.11940 using an ex- Seamount.Two lavas that are transitional in compositionbe- ponential mass-fractionationlaw [Wasserburg et al., 1981]. tween alkali basalt and tholeiite have 87Sr/86Srratios of During this study, 35 analyses of the standard SRM 987 0.7022 and 0.7023 ñ 0.0005, typical of a depleted source.In SRC03 (National Bureau of Standards) yielded a mean contrast,an alkali basalthas a higherratio of 0.7033 ñ 0.0002, 87Sr/86Srratio of 0.71023ñ 0.00004(2-a). The pooledstan- indicative of an enrichedsource. In thesesamples, unlike the dard deviationsfrom five replicate analysesprovide an esti- samplesalong the ridge, 8•Sr/86Sr ratios increase with increas- mate of the analyticalprecision of ñ0.00006 (2-a). More than ing K20/K20 + Na20. one sample was analyzed from eight dredge hauls, and the within-dredgemean 2-a standard deviation is _+0.00008;this DISCUSSION OF RESULTS suggeststhat intradredgevariability is not significantlygreater The geochemicalbehavior of the Juan de Fuca Ridge/Cobb than analyticalprecision. Hotspot systemis unlike that of other well-studiedridge/hot- The six microcrystalline-basaltsamples, though visually spot systems. The differences are apparent on plots of unalteredin thin section,were subjectto acid leachingin 10 8•Sr/86Srversus distance from the hotspot(Figure 6) for Ice- mL of 2.5 N HC1 for 10 min. Sr-isotopiccomposition was land, the Azores, the Galapagos spreading center, and the measuredon both leachedand unleachedportions; results are Juan de Fuca Ridge. Hotspots other than the Cobb Hotspot shownin Figure4. Only one leachedsample has a 8•Sr/86Sr have distinct maxima of Sr isotopic ratios (> 0.7030) that de- ratio more than 2-a lower than its unleachedpair. Though the crease systematicallytoward ratios more typical of normal leaching procedure was not as severe as that describedby MORB (0.7024-0.7030)with increasingdistance from the hot- O'Nions and Pankhurst [1976], the close agreement between spot. Schilling[1982a] and Verma and Schilling[1982] argued analysesof glassand powdersof microcrystalline-basaltsam- that the processesof fractional crystallization, partial melting, ples and the very small decreasein 8:Sr/S6Srratios upon or disequilibrium partial melting are inadequate to explain leachingsuggest that seawateralteration has not modified the these variations along the Galapagos spreading center, and 8:Sr/S6Srratios of the microcrystalline-basaltsamples signifi- proposed a binary mixing model between normal and en- cantly. riched MORB end-members. In contrast,the geochemicaleffect of the Cobb Hotspot on ANALYTICAL RESULTS the Sr isotopic ratio of lavas from the Juan de Fuca Ridge is The measured8:Sr/S6Sr ratios and samplelocations are undetectable; there is no systematic spatial variation of listed in Table 2 and plotted againstlatitude in Figure 5. In 87Sr/86Srratios. By conventionalmodels, the 87Sr/86Srratios general,the 87Sr/a6Srratios are at the lowerend of the range of lavas along the ridge suggestthat thesemagmas are derived for normal MORB and have an average value of from an isotopicallyhomogeneous depleted mantle. 7886 EABYET AL.' $R ISOTOPICRATIOS ALONG JUAN DE FUCARIDGE

o

o

½

o EABY ET AL.' $R ISOTOPIC RATIOS ALONG JUAN DE FUCA RIDGE 7887

the relation betweenbasalt chemistryand morphotectonicset- Effect of leaching ting is more complex than previously thought. Bryan et al. whole-rock powders in HCI [1976] report that enriched MORB (their type II) was recov- ered at ridge segmentsand Deep Sea Drilling Project (DSDP) drill sites not associatedwith hotspot-derivedoceanic islands or topographic highs. Results from two sites drilled by the

0.70240 International Program of Ocean Drilling (IPOD) Leg 82, de- signedto test temporal variationsof the AzoresHotspot, show ,/ that both depletedand enrichedMORB occur as intercalated lava flows [Bougault et al., 1983; Weaver and Tarney, 1983]. LeRoex et al. [1983] report that depleted and enriched MORB occur juxtaposed, even within a single dredge, throughout the region, longitude løW to longitude 11øE, / along the Southwest Indian Ridge, an extremely slowly / spreadingridge. They also find no evidencefor a geochemical / 0.70200 I I I I I I I , gradient away from the Bouvet Hotspot. The spatial and tem- 0.7021 0.7023 0.7025 0.7027 poral juxtaposition of enrichedand depletedM ORB suggests 87Sr/86Sr not leached that the mantle sourcesare heterogeneouson the scale of Fig. 4. Plot of 87Sr/86Srratios for leachedversus not leached kilometers to tens of kilometers. The absenceof geochemical microcrystalline-basaltsamples. Dashed line representsline of equal- ity. Only one leachedsample has an 87Sr/86Srratio more than 2-a gradientsaway from the Cobb and Bouvet hotspotssuggests lower than its unleachedpair. that the high volcanic productivity may be causedby thermal anomalies in the mantle rather than by enriched mantle plumes. Trace elementdata [Liias and Rhodes,1982; Schilling et al., Results from studies of small seamounts are consistent with 1982b] do not show a systematicgradient away from the a model of ubiquitoussmall heterogeneitiesin the mantle.In a Cobb Hotspot, thus supportingthe conclusionthat the hot- study of relatively small, isolated,oceanic central volcanoeson spot is nonenriched in nature. In contrast to the uniform Sr young crust near the East Pacific Rise, Batiza [1980] showed isotopic ratios, however, incompatible element ratios are that most volcanoesare composedof basalt chemicallysimilar highlyvariable, both along the wholeridge and withinsingle- to MORB. Some young volcanoescontain both depleted and dredge hauls; most are intermediate between normal and en- enriched basalt; compositional diversity seems to increase riched MORB [Liias and Rhodes, 1982]. The rare earth ele- with age [Batiza and Vanko, 1984]. Seamountsin ment patterns range from light rare earth elementdepleted to the northeastPacific including Denson, Davidson, and Hodg- light rare earth element enriched with a dominance of fairly kins seamounts of the Pratt-Welker chain [Turner et al., flat rare earth elementpatterns. :,ot,_•and the Heck •11tl---' Heckleseamount chains [Bart, i974] The higher87Sr/86Sr ratios of lavasfrom the southernend are interpretedto have eruptedon or near a spreadingcenter of region 1 could be due to assimilation of hydrothermally and are similar in major and minor elementcompositions to altered crust during protracted crystal fractionation [O'Hara, normal MORB. No Sr isotopic data for theseseamounts are 1977] or to subtle mantle heterogeneities.The lack of corre- availablefor comparison.The S?Sr/aaSrratios from Cobband lation betweenFeO* and 87Sr/S6Srratios for lavasalong the Brown Bear seamountsand the axial seamount suggestthat ridge argues against any simple relation between Sr isotopic central volcanoeson or near the Juan de Fuca Ridge are ratios and assimilationduring crystal fractionation. constructedof lavas derivedfrom depletedsources. The Cobb- The absenceof any variation in Sr isotopic ratios of lavas Eickelberg seamountchain is intermediate in size between the from regions4 and 3 is enigmaticbecause minor elementgeo- small volcanoes studied by Batiza and the large hotspot- chemistrysuggests that the lavas were derived from chemically generatedislands of Hawaii,Iceland, and Galapagos.The oc- distinctmantle sources(J. R. Delaney et al., unpublisheddata, currenceof both depleted and enriched basalt on Cobb Sea- 1984). This apparent decoupling could be due to mantle mount indicatesthat enrichedmagma may be tapped as the sourcesthat have not been separatedlong enough to develop volcanomoves farther away from the ridge axis. isotopicdifferences. Simple heterogeneousmantle models, however, cannot ex- plain the dominance of depleted compositionsproduced at IMPLICATIONS FOR MANTLE STRUCTURE mid-oceanic ridge spreading centers; production of basaltic Debate on the geometry of the chemically defined mantle melt from a mantle source with ubiquitous small heteroge- components is centered around two types of models. In neities should result in a wide range of compositionsdistrib- layered-mantlemodels [Schilling, 1973; Sun and Hanson,1975; uted randomly along mid-oceanicridges. Cohen and O'Nions Wasserburgand DePaolo, 1979] a relatively undifferentiated [1982] and Allegre et al. [1983] suggestthat spreadingrate is lower mantle (enriched source)is overlain by a more differ- an important variable controllingthe extent or compositional entiated (depleted source).The enriched lower variability of lava along a given spreadingcenter. Homogen- mantle can be introducedinto the upper mantle in the form of ization of melt compositionswill occur at fast spreading a rising plume (hotspot). In heterogeneousmantle models ,such as the East Pacific Rise, becausemultiple batches [Davis, 1981; Zindler, 1982; Sleep, 1984] the enriched hetero- of magma will mix in extensivemagma chambers.Transient geneitiesare small and ubiquitously distributed throughout magma chambers along slow spreading ridges, such as the the mantle. Both models implicitly assume that oceanic is- Southwest Indian and Mid-Atlantic Ridges, allow for the lands and topographichighs astride ridges are geneticallyre- eruption of singlebatches of magma with compositionsmore lated to the compositionof oceanicbasalt. representativeof mantle sources.The small range of Sr iso- Continued sampling of mid-oceanicridges is showing that topic ratios of basalt from the Juan de Fuca Ridge may be 7888 EABY ET AL.' $R ISOTOPIC RATIOS ALONG JUAN DE FUCA RIDGE

TABLE 2. Sr IsotopeAnalyses of BasaltFrom the Juande Fuca Ridge

Sample Latitude, Longitude, Depth, Number Source* RegionJ' deg north deg west m 87Sr/86Sr

TT152-37 1 1 44ø37.1 ' 130023.4 ' 2235-2240 0.70261 TT152-37-11 1 1 44ø37.1 ' 130023.4 ' 2235-2240 0.70255 L1181WF-5-64 4 1 44038.37' 130022.60 ' 2208 0.70259 L1181WF-5-67 4 1 44038.37 ' 130022.60 ' 2208 0.70267 L1181WF-17-18 4 1 44039.33' 130021.98 ' 2231 0.70256 Ll181WF-26-5 4 1 44042.07 ' 130020.72 ' 2228 0.70257 TT 152-43-19 1 1 44 ø59.6' 130 ø12.3' 2282 0.70263 Y74-1-4-8 3 1 45 ø16' 130008 ' 2400 0.70241 TT152-47-39 2 1 45017.7 ' 130005.2 ' 2400-2450 0.70254 Y74-1-3-33 3 1 45027 ' 130008 ' 1250 0.70253 Y 74-1- 3-41 3 1 45 o27' 130008 ' 1250 0.70256 TT152-77-6 1 2 45040.3 ' 130003.8 ' 1920-2060 0.70244 TT152-77-7 1 2 45040.3 ' 130003.8 ' 1920-2060 0.70252 TT152-50-25 2 2 45042.5 ' 130002.5 ' 2100 0.70256 Y74-1-1-7 3 2 45 ø46' 130001' 2169 0.70245 Y74-1-1-34 3 2 45ø46 ' 130001 ' 2169 0.70252 TT152-55-25 1 2 45056.3 ' 129059.6 ' 1530-1557 0.70257 TT170-5 1 2 45056.9 ' 130001.8 1425-1560 0.70246 TT152-53-6 2 2 45057.7 ' 130003.3 ' 1460-1510 0.70250 TT152-61-1 1 2 46004.5 ' 129058.08 ' 2215-2355 0.70249 TT152-72 1 2 46016.7 ' 129043.8 ' 2265-2366 0.70237 TT152-65-56 2 2 46021.2 ' 129040.7 ' 2295-2320 0.70233 TT152-33-3 1 3 46039.2 ' 129025.6 ' 2425-2511 0.70245 TT152-29-1 1 3 46051.6 ' 129017.9 ' 2310-2360 0.70243 TT152-21 1 3 46055.5 ' 129015.7 ' 2410-2520 0.70248 TT152-21-41 2 3 46055.5 ' 129015.7 ' 2410-2520 0.70246 TT 152-13 1 3 47 ø12.5' 129007.2 ' 2625 0.70243 HU 81-017-6-11 1 3 47 ø12.6' 129004.87 ' 2600 0.70245 TT 152-6-81 2 3 47 ø16.1' 129004.1' 2575 9.70240 TT152-11-21 1 3 47032.5 ' 128057.8 ' 2640-2690 0.70247 HU81-017-11 1 3 47ø37.21 ' 129ø17.31 ' 1390 0.70254 TT170-72-2 1 4 47042.49 ' 129016.55 ' 2700 0.70259 HU81-017-2-4 1 4 47 ø43.6' 129 ø14.8' 2600 0.70241 TT152-1-17 2 4 47050.7 ' 129009.3 ' 2440 0.70243 HU81-017-20-11 1 4 47 ø59.9' 128 ø59.4' 2125 0.70250 HU 81-017-20-21 1 4 47 ø59.9' 128 ø59.4' 2125 0.70243 HU81-017-20-27 1 4 47 ø59.9' 128 ø59.4' 2125 0.70243 HU81-017-17-4 1 4 48 ø12.0' 128 ø58.9' 2615 0.70250 HU 81-017-31-2 1 4 49003.3 ' 130 ø56.7' 1200 0.70241 TT170-20-2 1 BBS 45059.76 ' 130025.73 ' 2100 0.70251

Sample HU81-017-31-2 is from Explorer Seamount and is not shown on the figures. Estimated analyticalprecision for 87Sr/86Sranalyses is +0.00006. *Source codes are 1, UW glass' 2, UW microcrystalline-basalt;3, OSU glass' 4, U.S. Geological Surveymicrocrystalline-basalt. 9Region codes are 1, latitude 44ø30'-45ø28'Nß 2, latitude 45ø28'-46ø39'Nß 3, latitude 46ø39'-47ø40'N; 4, > latitude 47ø40'N ßBBS, Brown Bear Seamount.

0.70270

ß 0.70260 ß

ß ß e ß ß ß

ß ß ß 0.70250 ß ß

ß ß ß ß ß ß ß

ß 0.70240 ß

ß 0.70230 a b c •'•

COBB REGION 1 REGION 2 REGION 3 OFFSET REGION 4 , I , I , I I , 45 46 47 48 LATITUDE (ß N) Fig. 5. Regionalvariation of 87Sr/86Srratios. A sample2-• errorbar is shownon the right-handside of the figure.The Sr isotopicdata do not showsystematic variation relative to the Cobb Hotspot. EABY ET AL.' SR ISOTOPIC RATIOS ALONG JUAN DE FUCA RIDGE 7889

TABLE 3. Data Used for Calculationof the Probabilityof would give rise to depletedMORB upon partial melting at Statistical Difference shallow levels. Group Location Mean S7Sr/S6Sr Sleep'smodel also addressesthe dependenceof large ion lithophileenrichment of hotspotmagmas on (1) the sizeof the a Southerntip of region 1 0.70260 _ 9 heterogeneities,(2) the total volumeof enrichedmaterial orig- b Spatially bracketing the 0.70252 _ 14 inally presentin the mantle,and (3) the efficienciesof melting axial seamount c Lowest ratios' northern 0.70235 _ 6 and melt removalduring ascent of a mantlediapir. The source flank of axial seamount regionfor the Juan de Fuca Ridge and the Cobb Hotspot, d Regions3 and 4 0.70246 _+9 according to Sleep'smodel, would be almost completely stripped of enriched melts before reaching shallow levels. For the data pair of a-b, probability is 0.47' for a-c, 0.98' for a-d, Small volumesof enrichedmelt are more likely to be tapped 0.73' for b-c, 0.88' for b-d, 0.36' and for c-d, 0.69. as seamountsmove off axis than at normal ridge segments. Tapping of melts from somewhatgreater depths at slow spreadingridges than at fast ridges, or diversion of enriched caused by averaging of heterogeneousmagma batches by magma back into the ridge along fracturesare two possible mixing in long-livedmagma chambers.Mixing of relatively explanationsfor the chemical enrichmentsexhibited by the undifferentiatedmagmas and residualdifferentiates in long- other hotspotssuch as Iceland,the Azores,and the Galapagos lived magmachambers that are periodicallyreplenished is in- spreadingcenter. dicated by mineral compositionaldata [Clague et al., 1983]. The Juan de Fuca data imply that the depletedmantle source CONCLUSIONS end-memberhas a S7Sr/S6Srratio lessthan 0.70235. A systematicSr isotopic gradient does not exist along the Sleep[1984-1 has proposeda model usingmantle convection Juan de Fuca Ridge away from the Cobb Hotspot. There is no dynamicsin a heterogeneousmantle that offersan explanation evidenceto support a model in which an enriched mantle for the formation of depletedmantle sourcesfor MORB and plumeis responsiblefor the zone of high volcanicproductivity enrichedmantle sourcesfarther away from the spreadingaxis. building the axial seamount.The small range in Sr isotopic Sleepargues that the directionof shearstress in the upwelling ratios along the Juan de Fuca Ridge may be related to hom- mantle diapir is such that the axis of tensionplunges away ogenizationby mixing in long-lived magma chambersas sug- from the ridge at great depthsand is nearly verticalat shallow gestedby Cohenand O'Nions [1982] or Allegre et al. [1983] levels.Enriched heterogeneities melt preferentiallyin compari- or to lateral migration away from the ridge axis of enriched son with the surroundingmantle. At great depths,these early meltsgenerated at greaterdepths as proposedby Sleep[ 1984]. enrichedmelts enter fracturesthat carry them away from the Sr isotopic ratios of samplesnorth and south of the Cobb ridge axis.The residualmantle diapir would be depletedrela- offset are identical. The decouplingof minor element and iso- tive to the averagemantle, owing to melting eventsearly in topic compositionssuggests that the sourceregion heteroge- the ,,a• ...... •*•"ø •'; s•,,.* .... • and to ' .... of t.•" .... •a•y '-' enrichedmeltS,...... and neitiesformed recently.

300 200 100 0 100 200 300 400 300 200 100 0 100 200 300 400 I I I I I I ! I I ! I I I I

_ s _ N S _ ß I _

ß

ß

ß ß ß ./'

JUAN DE FUCA - AZORES 0.70280 03 (White et al, 1976) ,•.

_ w s _ _- -_ Ridge

ß 0.70300 ß 0.70280

ICELAND ß ß

(Hart el al. 1973) ß o.7o2no0.70260f(v GA'LAPAGO$ I I I I I I I I I I { 300 200 100 0 100 200 300 400 300 200 100 0 100 200 300 400 Distance (kin) from hot spot Distance (km) from hot spot Fig. 6. 87Sr/86Srratios versus distance from hotspot for Iceland[Hart et al., 1973],the Azores[White et al., 1976],the Galapagosspreading center [Verrna and Schilling,1982], and Juan de Fuca Ridge.Unlike Sr isotopicratios from Juan de Fuca Ridge, Sr isotopicratios from Iceland, Azores,and Galapagosspreading center have distinct maxima (>0.7030) at the hotspot that decreasesystematically toward ratios more typical of normal MORB with increasingdistance from the hotspot. 7890 EABY ET AL.: SR ISOTOPICRATIOS ALONG JUAN DE FUCA RIDGE

Acknowledgments.We expressour appreciationto Debbie Kelley O'Nions, R. K., and R. J. Pankhurst, Sr isotope and rare earth ele- for hand picking the glasssamples, Jack Dymond for supplyingsam- ment geochemistryof DSDP Leg 37 , Earth Planet. Sci. plesfrom OSU, and Anne Tapay for operatingthe massspectrometer. Lett., 31, 255-261, 1976. Thorough reviews by Marvin Lanphere, Robert Christiansen,Jean- O'Nions, R. K., P. J. Hamilton, and N.M. Evensen, Variations in Guy Schilling, and Rodey Batiza greatly improved the manuscript. •½3Nd/•½½Ndand 87Sr/86Srratios in oceanicbasalts, Earth Planet. Some of the ideas are the result of discussionswith Norm Sleep, Gail Sci. Lett., 34, 13-22, 1977. Mahood, Clark Johnson,Jan Morton, John Lupton, Roger Hart, and Schilling, J.-G., Iceland geochemicalevidence along Jack Dymond. This project was supported in part by NSF grant ReykjanesRidge, Nature, 242, 565-571, 1973. OCE81-11413. Schilling,J.-G., Rare-earth variations across"normal" segementsof the ReykjanesRidge, 60ø-53øN,Mid-Atlantic Ridge,29øS, and East REFERENCES Pacific Rise, 2ø-19øS,and evidenceon the compositionof the un- Allegre, C. J., B. Dupre, and B. Hamelin, Geochemistryof oceanic derlying low velocity layer, J. Geophys.Res., 80, 1459-1473, 1975a. Schilling, J.-G., Azores mantle blob: Rare earth evidence, Earth ridge basalts explained by blob mantle mixing, Eos Trans. A GU, Plant. Sci. Lett., 25, 103-115, 1975b. 64, 324, 1983. Barr, S. M., Seamount chains formed near the crest of the Juan de Schilling, J.-G., R. Kingsley, and J. Devine, Galapagos hot spot- Fuca Ridge,northeast Pacific Ocean, Mar. Geol.,17, 1-19, 1974. spreadingcenter system, 1, Spatial petrological and geochemical Batiza, R., Origin and petrology of young oceaniccentral volcanoes: variations(83øW-101øW), J. Geophys.Res., 87, 5593-5610, 1982a. Are most tholeiitic rather than alkalic?, Geology,8, 477-482, 1980. Schilling,J.-G., B. Cousens,R. L. Chase, and J. R. Delaney, Juan de Batiza, R., and D. Vanko, Petrology of young Pacific seamounts,or. Fuca-Explorer Ridge: Rare earth contents,Eos Trans. AGU, 63, 1154, 1982b. Geophys.Res., in press,1984. Sinton, J. M., D. S. Wilson, D. M. Christie, R. N. Hey, and J. R. Bougault, H., S. Cande, J. L. Joron, and M. Treuil, Local versus Delaney, Petrologic consequencesof rift propagation on oceanic regional mantle heterogeneities:IPOD Leg 82 of the Glomar Challenger,Eos Trans. AGU, 64, 345, 1983. spreadingridges, Earth Planet. Sci. Lett., 62, 193-207, 1983. Sleep, N.H., Tapping of from ubiquitous mantle heteroge- Bryan, W. B., G. Thompson, F. A. Frey, and J. S. Dickey, Inferred neities,J. Geophys.Res., in press,1984. geologicsettings and differentiationin basaltsfrom the Deep-Sea Subbarao,K. V., G. S. Clark, and R. B. Forbes, Strontium isotopesin Drilling Project, J. Geophys.Res., 81, 4285-4304, 1976. some seamount basalts from the northeastern Pacific ocean, Can. J. Clague, D. A., and T. E. Bunch, Formation of ferrobasalt at East Earth Sci., 10, 1479, 1973. Pacific midocean spreading centers, J. Geophys.Res., 81, 4247- Sun, S.S., and G. N. Hanson, Evolution of the mantle: Geochemical 4256, 1976. • evidencefrom alkali basalt, Geology,3, 297-302, 1975. Clague, D. A., F. A. Frey, G. Thompson, and S. Rindge, Minor and Turner, D. L., R. D. Jarrard, and R. B. Forbes, Geochronology and trace elementgeochemistry of volcanicrocks dredgedfrom the Ga- origin of the Pratt-Welker SeamountChain, Gulf of Alaska: A new lapagos spreadingcenter: Role of crystal fractionation and mantle pole of rotation for the , J. Geophys.Res., 85, 6547- heterogeneity,J. Geophys.Res., 86, 9469-9482, 1981. 6556, 1980. Clague, D. A., J. S. Eaby, and J. R. Delaney, Petrologicevidence for Verma, S. P., and J.-G. Schilling, Galapagos hot spot-spreading long-lived axial magma chamberson the Juan de Fuca Ridge (ab- centersystem, 2, 87Sr/86Srand large ion lithophileelement vari- stract),paper presentedat the Fifth Annual GeodynamicsResearch ations (85øW-101øW), J. Geophys.Res., 87, 10838-10856, 1982. Symposium on Oceanic Lithosphere: Origin, Structure, and Dy- Vogt, P. R., and G. R. Byerly, Magnetic anomaliesand basalt compo- namics,Texas A&M University, 1983. sition in the Juan de Fuca-Gorda Ridge area, Earth Planet. Sci. Cohen, R. S., and R. K. O'Nions, The lead, neodymium and stron- Lett., 33, 185-207, 1976. tium isotopic structure of ocean ridge basalts,J. Petrol., 23, 299- 324, 1982. Vogt, P. R., and G. L. Johnson,Transform faults and longitudinal flow below the midoceanicridge, J. Geophys.Res., 80, 1399-1428, Davies, G. F., Earth's neodymiumbudget and structureand evolution 1975. of the mantle, Nature, 290, 208, 1981. Delaney, J. R., H. P. Johnson, and J. L. Karsten, The Juan de Fuca Wasserburg,G. J., and D. J. DePaolo, Models of earth structure inferred from neodymium and strontium isotopic abundances, Ridge-hotspot-propagatingrift system:New tectonic,geochemical, Proc. Nat. Acad. Sci., 76(8), 3594-3598, 1979. and magneticdata, J. Geophys.Res., 86, 11747-11750, 1981. Delaney, J. R., J. L. Karsten, and D. A. Clague, Geochemistryof Juan Wasserburg,G. J., S. B. Jacobsen,D. J. DePaolo, M. T. McCulloch, and T. Wen, Precise determination of Sm/Nd ratios, Sm and Nd de Fuca Ridge basaltic glasses,Eos Trans. AGU, 63, 1147, 1982. isotopic abundance in standard solution, Geochim.Cosmochim. DePaolo, D. J., and G. J. Wasserburg, Inferences about magma sourcesand mantlestructure from variationsof •'•3Nd/•'•'•Nd,Geo- Acta, 45, 2311-2324, 1981. Weaver, B. L., and J. Tarney, Trace element geochemistryof basalts phys.Res. Lett., 3, 743-746, 1976. recoveredby DSDP Leg 82: Implications for mantle heterogeneity, Hart, S. R., J.-G. Schilling,and J. L. Powell, Basaltsfrom Iceland and Eos Trans. AGU, 64, 345, 1983. along the ReykjanesRidge: Sr isotope geochemistry,Nature Phys. White, W. M., and W. B. Bryan, Sr-isotope, K, Rb, Cs, Sr, Ba, and Sci., 246(155), 104-107, 1973. rare-earth geochemistryof basaltsfrom the FAMOUS area, Geol. Hedge, (2. E., and Z. E. Peterman, The strontium isotope composition Soc. Am. Bull., 88, 571-576, 1977. of basalts from the Gordo and Juan de Fuca Rises, northeastern White, W. M., and J.-G. Schilling,The nature and origin of geochemi- Pacific Ocean, Contrib. Mineral. Petrol., 27, 114-120, 1970. cal variations in Mid-Atlantic Ridge basaltsfrom the central North Hey, R. N., and D. S. Wilson, Propagating rift explanation for the Atlantic, Geochim. Cosmochim.Acta., 42, 1501-1516, 1978. tectonic evolution of the northeast Pacific--The pseudomovie, Earth Planet. Sci. Lett., 58, 167-188, 1982. White, W. M., S. R. Hart, and J.-G. Schilling,Geochemistry of the Azores and the Mid-atlantic Ridge: 29øN to 60øN, Carnegie Inst. Johnson, H. P., J. L. Karsten, J. R. Delaney, E. E. Davies, R. G. Wash. Yearb., 74, 224-234, 1975. (2urrie, and R. L. Chase,A detailed study of the Cobb Offset of the White, W. M., J.-G. Schilling,and S. R. Hart, Evidencefor the Azores Juan de Fuca Ridge: Evolution of a propagating rift, J. Geophys. Res., 88, 2297-2315, 1983. mantle plume from strontium isotope geochemistryof the central north Atlantic, Nature, 263, 659-663, 1976. Kay, R., N.J. Hubbard, and P. W. Gast, Chemical characteristicsand Wilson, J. T., Evidence from ocean islands suggestingmovement in origin of oceanic ridge volcanic rocks, J. Geophys.Res., 75, 1585- the earth, Phil. Trans. R. Soc. London, 258, 145-165, 1965. 1613, 1970. LeRoex,A. P., H. J. B. Dick, A. J. Erlank, A.M. Reid, F. A. Frey, and Zindler, A., E. Jagoutz, and S. Goldstein, Nd, Sr and Pb isotopic S. R. Hart, Geochemistry,mineralogy and petrogenesisof lavas systematicsin a three-component mantle: A new perspective, eruptedalong the southwestIndian ridgebetween the Bouvettriple Nature, 298, 519, 1982. junction and 11 degreesEast, J. Petrol.,24, 267-318, 1983. Liias,R. A., and J. M. Rhodes,Evolution of basalticmagma along the D. A. Clague and J. Eaby, U.S. Geological Survey, MS 999, 345 Juan de Fuca Ridge, Eos Trans. AGU, 63, 1147, 1982. Middlefield Road, Menlo Park, CA 94025. Morgan, W. J., Plate motions and deep mantle convection,in Studies J. R. Delaney, OceanographyDepartment, WB-10, University of in Earth and SpaceSciences, Hess Volume,Mem. 132, editedby R. Washington,Seattle, WA 98195. Shaganet al., GeologicalSociety of America,Boulder, Colo., 1973. O'Hara, M. J., Geochemicalevolution during fractional crystallis- (ReceivedNovember 11, 1983; ation of a periodically refilled magma chamber, Nature, 266, 503- revisedApril 30, 1984; 508, 1977. acceptedMay 8, 1984.)